bogoreader.c

一个C语言写的快速贝叶斯垃圾邮件过滤工具

static bool mailbox_next_mail(void)
{
    bool val = mail_first || have_message;
    mail_first = false;
    return val;
}

/* iterates over files in a directory */
static bool dir_next_mail(void)
{
    struct dirent *dirent;
    struct stat st;

    for (;;) {
	if (reader_dir == NULL) {
	    char *x = dir_name;
	    /* open next directory */
	    if (mailstore_type == MS_MAILDIR) {
		size_t siz;

		if (*maildir_sub == NULL)
		    return false; /* IMPORTANT for termination */
		siz = strlen(dir_name) + 4 + 1;
		x = xmalloc(siz);
		strlcpy(x, dir_name, siz);
		strlcat(x, *(maildir_sub++), siz);
	    }
	    reader_dir = opendir(x);
	    if (!reader_dir) {
		fprintf(stderr, "cannot open directory '%s': %s", x,
			strerror(errno));
	    }
	    if (x != dir_name)
		xfree(x);
	}

	while ((errno = 0, dirent = readdir(reader_dir)) != NULL) {
	    /* skip private files */
	    if ((mailstore_type == MS_MAILDIR && dirent->d_name[0] != '.') ||
		(mailstore_type == MS_MH && isdigit((unsigned char)dirent->d_name[0])))
		break;
	}

	if (errno) {
	    fprintf(stderr, "Cannot read directory %s: %s",
		    dir_name, strerror(errno));
	    exit(EX_ERROR);
	}

	if (dirent == NULL) {
	    if (reader_dir)
		closedir(reader_dir);
	    reader_dir = NULL;
	    if (mailstore_type == MS_MAILDIR)
		continue;
	    if (mailstore_type == MS_MH)
		return false;
	}

	filename = namebuff;
	snprintf(namebuff, sizeof(namebuff), "%s%s%c%s", dir_name, 
		 (mailstore_type == MS_MH) ? "" : *(maildir_sub-1),
		 DIRSEP_C, dirent->d_name);

	bogoreader_close();
	fpin = fopen( filename, "r" );
	if (fpin == NULL) {
	    fprintf(stderr, "Warning: can't open file '%s': %s\n", filename,
		    strerror(errno));
	    /* don't barf, the file may have been changed by another MUA,
	     * or a directory that just doesn't belong there, just skip it */
	    continue;
	}

	/* skip non-regular files */
	if (0 == fstat(fileno(fpin), &st) && !S_ISREG(st.st_mode))
	    continue;

	if (DEBUG_READER(0))
	    fprintf(dbgout, "%s:%d - reading %s (%p)\n", __FILE__, __LINE__,
		    filename, (void *)fpin);

	return true;
    }
}

/*** _getline functions ***********************************************/

/* reads from a mailbox, paying attention to ^From lines */
static int mailbox_getline(buff_t *buff)
{
    uint used = buff->t.leng;
    byte *buf = buff->t.text + used;
    int count;
    static word_t *saved = NULL;
    static bool emptyline = false;		/* for mailbox /^From / match */

    if (saved != NULL) {
	count = saved->leng;
	buff_add(buff, saved);
	word_free(saved);
	saved = NULL;
	return count;
    }

    count = buff_fgetsl(buff, fpin);
    have_message = false;

    /* XXX FIXME: do we need to unescape the >From, >>From, >>>From, ... lines
     * by discarding the first ">"? */

    /* DR 08/25/03 - NO!!! */

    if ((firstline || emptyline) &&
	seplen != 0 && count >= (int) seplen && memcmp(separator, buf, seplen) == 0)
    {
	if (firstline) {
	    firstline = false;
	}
	else {
	    have_message = true;
	    saved = word_new(buf, count);
	    count = EOF;
	}
    }
    else {
	if (buff->t.leng < buff->size)		/* for easier debugging - removable */
	    Z(buff->t.text[buff->t.leng]);	/* for easier debugging - removable */
    }

    emptyline = is_eol((char *)buf, count);

    return count;
}

/* reads from an rmail batch, paying attention to ^#! rmail lines */
static int rmail_getline(buff_t *buff)
{
    int count;
    uint used = buff->t.leng;
    byte *buf = buff->t.text + used;
    static word_t *saved = NULL;
    static unsigned long bytesleft = 0;

    if (saved != NULL) {
	count = saved->leng;
	buff_add(buff, saved);
	word_free(saved);
	saved = NULL;
	return count;
    }

    if (bytesleft) {
	count = buff_fgetsln(buff, fpin, bytesleft);
	bytesleft -= count;
	return count;
    }

    count = buff_fgetsl(buff, fpin);
    have_message = false;

    if (count >= (int) seplen && memcmp(separator, buf, seplen) == 0)
    {
	uint i;
	bytesleft = 0;
	for (i = seplen; i < (uint) count; i++) {
	    if (isspace(buf[i])) continue;
	    if (!isdigit(buf[i])) break;
	    bytesleft = bytesleft * 10 + (buf[i] - '0');
	}
	if (firstline) {
	    firstline = false;
	}
	else {
	    have_message = true;
	    saved = word_new(buf, count);
	    count = EOF;
	}
    } else {
	if (buff->t.leng < buff->size)		/* for easier debugging - removable */
	    Z(buff->t.text[buff->t.leng]);	/* for easier debugging - removable */
    }

    return count;
}

/* reads from an ANT batch, paying attention to ^#! rmail lines */
static int ant_getline(buff_t *buff)
{
    int count;
    uint used = buff->t.leng;
    byte *buf = buff->t.text + used;
    static word_t *saved = NULL;
    static bool dot_found = true;

    if (saved != NULL) {
	count = saved->leng;
	buff_add(buff, saved);
	word_free(saved);
	saved = NULL;
	return count;
    }

    count = buff_fgetsl(buff, fpin);
    have_message = false;

    if (dot_found && count >= (int) seplen && memcmp(separator, buf, seplen) == 0)
    {
	dot_found = false;		/* ignore until dot */
	if (firstline) {
	    firstline = false;
	}
	else {
	    have_message = true;
	    saved = word_new(buf, count);
	    count = EOF;
	}
    } else {
        if ((count == 2 || count == 3) && 
	    (buf[0] == '.') && 
	    (buf[1] == '\r' || buf[1] == '\n'))
            dot_found = true;			/* dot found.  look for separator */
	if (buff->t.leng < buff->size)		/* for easier debugging - removable */
	    Z(buff->t.text[buff->t.leng]);	/* for easier debugging - removable */
    }

    return count;
}

/* reads a file as a single mail ( no ^From detection ). */
static int simple_getline(buff_t *buff)
{
    int count = buff_fgetsl(buff, fpin);

    if (buff->t.leng < buff->size)	/* for easier debugging - removable */
	Z(buff->t.text[buff->t.leng]);	/* for easier debugging - removable */

    return count;
}

/* initialize for MH directory and
 * Maildir subdirectories (cur and new). */
static void dir_init(const char *name)
{
    reader_dir = NULL;
    fini = dir_fini;

    if (mailstore_type == MS_MAILDIR)
	maildir_sub = maildir_subs;

    save_dirname(name);

    return;
}

/* finish up MH/Maildir store */

static void dir_fini(void)
{
    if (reader_dir)
	closedir(reader_dir);
    reader_dir = NULL;
    return;
}

/* returns current file name */
static const char *get_filename(void)
{
    return filename;
}

/* global reader initialization, exported */
void bogoreader_init(int _argc, char **_argv)
{
    mailstore_first = mail_first = true;
    reader_more = reader__next_mail;
    fini = dummy_fini;
    switch (bulk_mode) {
    case B_NORMAL:		/* read mail (mbox) from stdin */
	yy_file = fpin;
	mailstore_next_store = stdin_next_mailstore;
	if (run_type & (REG_SPAM|REG_GOOD|UNREG_SPAM|UNREG_GOOD))
	    mbox_mode = true;
	break;
    case B_STDIN:		/* '-b' - streaming (stdin) mode */
	mailstore_next_store = b_stdin_next_mailstore;
	break;
    case B_CMDLINE:		 /* '-B' - command line mode */
	argc = _argc;
	argv = _argv;
	mailstore_next_store = b_args_next_mailstore;
	mailstore_next_mail  = NULL;
	break;
    default:
	fprintf(stderr, "Unknown bulk_mode = %d\n", (int) bulk_mode);
	abort();
	break;
    }
    reader_filename = get_filename;
}

/* For bogoconfig to distinguish '-I file' from '-I dir' */
/* global reader initialization, exported */
void bogoreader_name(const char *name)
{
    bool ok;
    switch (isdir(name)) {
    case IS_FILE:
	fpin = fopen( name, "r" );
	ok = fpin != NULL;
	break;
    case IS_DIR:
	ok = open_mailstore(name);
	break;
    default:
	ok = false;
	break;
    }
    if (!ok) {
	fprintf(stderr, "Can't read '%s'\n", name);
	exit(EX_ERROR);
    }
}

/* cleanup after reading a message, exported. */
/* Only called if the passthrough code says it is ok to close the file. */

void bogoreader_close_ifeof(void)
{
    if (fpin && feof(fpin))
       bogoreader_close();
}

/* global cleanup, exported */
void bogoreader_fini(void)
{
    bogoreader_close();
    fini();
}

static void bogoreader_close(void)
{
    if (fpin && fpin != stdin)
	fclose(fpin);
    fpin = NULL;
}